基于ABC-SVM算法的车牌识别系统设计
|
摘要
为了准确快速的识别车牌,设计并实现了一种基于人工蜂群算法优化支持向量机(ABC-SVM)的车牌识别系统。该系统以STM32单片机为主控,通过控制OV7725摄像头进行车牌图像采集,采用4G无线模块将采集到的车牌图像发送给上位机进行处理识别。上位机利用人工蜂群算法对支持向量机的惩罚系数和核参数进行优化,并结合支持向量机构建车牌字符识别模型。测试结果表明,系统能够准确快速的识别车牌,识别准确率(93.3%)较传统BP神经网络(85.0%)和支持向量机(88.1%)分别提高了8.3%和5.2%。
In order to accurately and quickly recognize the license plate,a license plate recognition system based on ABC-SVM algorithm was designed and implemented in this paper.In this license plate recognition system,the STM32 microcontroller is used as the main controller to control OV7725 camera to collect license plate images,and the collected images are send to the upper computer for further recognition processing through 4 Gwireless module.In the upper computer,the penalty coefficient and the kernel function parameter of support vector machine are optimized by artificial bee colony algorithm,and the license plate character recognition model is setup combined with support vector machine.The test results show that this system can accurately and quickly recognize the license plate,and its recognition accuracy can be increased by 8.3% and 5.2%,reaching 93.3%,while the conventional support vector machine and BP neural network can reach only 85.0% and 88.1%.
In order to accurately and quickly recognize the license plate,a license plate recognition system based on ABC-SVM algorithm was designed and implemented in this paper.In this license plate recognition system,the STM32 microcontroller is used as the main controller to control OV7725 camera to collect license plate images,and the collected images are send to the upper computer for further recognition processing through 4 Gwireless module.In the upper computer,the penalty coefficient and the kernel function parameter of support vector machine are optimized by artificial bee colony algorithm,and the license plate character recognition model is setup combined with support vector machine.The test results show that this system can accurately and quickly recognize the license plate,and its recognition accuracy can be increased by 8.3% and 5.2%,reaching 93.3%,while the conventional support vector machine and BP neural network can reach only 85.0% and 88.1%.
引文
[1]张松,王飞.嵌入式车牌识别系统的硬件电路设计[J].现代电子技术,2012,35(10):1-3.
[2]胡玉琛,刘一清.基于FPGA的车牌识别系统研究与实现[J].电子测量技术,2015,38(12):54-58.
[3]瞿国庆,李汪佩.渣土车车牌字符智能识别研究[J].电子测量与仪器学报,2016,30(12):1853-1860.
[4]郭荣艳,胡雪惠.BP神经网络在车牌字符识别中的应用研究[J].计算机仿真,2010,27(9):299-301.
[5]司朋举,胡伟.一种改进的神经网络车牌识别算法研究[J].电子测量技术,2016,39(10):100-103.
[6]孙晶晶,静大海.基于神经网络复杂背景下车牌识别系统的研究[J].国外电子测量技术,2016,36(8):22-29.
[7]王磊,王瀚漓,何良华.基于双边缘检测的车牌识别算法[J].计算机工程与应用,2013,49(8):169-173.
[8]张加宏,潘周光,刘进.基于ARM的远程室内环境监测与调控系统设计[J].现代电子技术,2017,40(2):12-15.
[9]刘爽.基于数学形态学的车牌定位识别系统设计[J].自动化与仪器仪表,2012,64(6):84-85.
[10]王永杰,裴明涛,贾云得.多信息融合的快速车牌定位[J].中国图象图形学报,2014,19(3):471-475.
[11]GHAHNAVIEH A E,AMIRKHANI-SHAHRAKIA,RAIE A A.Enhancing the license plates character recognition methods by means of SVM[C].22nd Iranian Conference on Electrical Engineering(ICEE),2014:220-225.
[12]行鸿彦,张强,徐伟.混沌海杂波背景下的微弱信号检测混合算法[J].物理学报,2015,64(4):52-58.
[13]XIAO X,HUANG X R,KANG Q.A Hill-climbingmethod-based maximum-power-point-tracking strategy for direct-drive wave energy converter[J].IEEETransactions on Industrial Electronics,2016,63(1):257-267.
[14]LU J,WANG H B,SUN G C.Frequency selective surface structure optimized by genetic algorithm[J].Chinese Physics B,2009,18(4):1598-1600.
[15]杨小东,刘波.人工蜂群算法加速收敛技术研究[J].计算机技术与发展,2014,24(4):25-28.
[16]吴一全,周杨,龙云淋.基于蜂群优化投影寻踪的高光谱小目标检测[J].仪器仪表学报,2016,37(6):1347-1355.
[2]胡玉琛,刘一清.基于FPGA的车牌识别系统研究与实现[J].电子测量技术,2015,38(12):54-58.
[3]瞿国庆,李汪佩.渣土车车牌字符智能识别研究[J].电子测量与仪器学报,2016,30(12):1853-1860.
[4]郭荣艳,胡雪惠.BP神经网络在车牌字符识别中的应用研究[J].计算机仿真,2010,27(9):299-301.
[5]司朋举,胡伟.一种改进的神经网络车牌识别算法研究[J].电子测量技术,2016,39(10):100-103.
[6]孙晶晶,静大海.基于神经网络复杂背景下车牌识别系统的研究[J].国外电子测量技术,2016,36(8):22-29.
[7]王磊,王瀚漓,何良华.基于双边缘检测的车牌识别算法[J].计算机工程与应用,2013,49(8):169-173.
[8]张加宏,潘周光,刘进.基于ARM的远程室内环境监测与调控系统设计[J].现代电子技术,2017,40(2):12-15.
[9]刘爽.基于数学形态学的车牌定位识别系统设计[J].自动化与仪器仪表,2012,64(6):84-85.
[10]王永杰,裴明涛,贾云得.多信息融合的快速车牌定位[J].中国图象图形学报,2014,19(3):471-475.
[11]GHAHNAVIEH A E,AMIRKHANI-SHAHRAKIA,RAIE A A.Enhancing the license plates character recognition methods by means of SVM[C].22nd Iranian Conference on Electrical Engineering(ICEE),2014:220-225.
[12]行鸿彦,张强,徐伟.混沌海杂波背景下的微弱信号检测混合算法[J].物理学报,2015,64(4):52-58.
[13]XIAO X,HUANG X R,KANG Q.A Hill-climbingmethod-based maximum-power-point-tracking strategy for direct-drive wave energy converter[J].IEEETransactions on Industrial Electronics,2016,63(1):257-267.
[14]LU J,WANG H B,SUN G C.Frequency selective surface structure optimized by genetic algorithm[J].Chinese Physics B,2009,18(4):1598-1600.
[15]杨小东,刘波.人工蜂群算法加速收敛技术研究[J].计算机技术与发展,2014,24(4):25-28.
[16]吴一全,周杨,龙云淋.基于蜂群优化投影寻踪的高光谱小目标检测[J].仪器仪表学报,2016,37(6):1347-1355.